Editor's note: As we wind down to the end of 2015, Constellation Research Insights is taking a look back at the year's highlights with perspective from each member of the analyst team. In this installment, we check in with Constellation Research VP and principal analyst Doug Henschen for his views on a big year in BI and analytics.

What was the most important market trend or trends in your coverage area during 2015?

Henschen: The theme that cut across several different areas was self-service. Five or 10 years ago it started with BI, but this year it seemed to be clicking in the areas of data access, data prep and advanced analytics, as well as continuing efforts to make reporting, querying and analysis more accessible. It fits in with consumerization of IT, and the overall idea of making software more accessible to untrained business users. It speaks to the whole notion of constrained resources and what people would rather do it for themselves rather than stand in a queue and have IT do it for them.

Another was the idea of making big data accessible to BI tools that lots of end users are familiar with, such as Tableau, Microstrategy and Cognos. So you saw a raft of emerging vendors, including ArcadiaData, AtScale, Paxata and Kyvos, working at making the data in Hadoop accessible. 

What was the biggest news story in your coverage area this year?

There was a lot of Apache Spark news in 2015, and the bigger reason for this is this idea there's not one silver bullet to take advantage of big data. That's why Spark was so much in the news. It represents the need for a variety, or ensemble of approaches—not just SQL, not just R, not just graph, not just streaming—all of those. 

A key driver is the fact that companies are generating and need to analyze a variety of data types in an increasingly digital world. As companies do more marketing and business online, clickstreams, social data and mobile data become much more important. SQL is good for analyzing the transactional data we've had for years, but graph and time-series analysis and machine learning and other techniques shine with these new types of data. They want answers more quickly, and that's driving interest in streaming and real-time approaches. And they want to look forward, not back, and do the right thing to maximize sales and profits, so that's driving demand for prediction, as you get from R. So there you have it: Companies need a variety of techniques —SQL, R, ML, Graph, streaming—to make the most of a variety of data.

Any predictions for 2016?

It's an easy prediction to make but we're going to see some serious consolidation in 2016. There are lots of startups in some of these self-service areas. Self-service is a feature set, not a company, so I think startups that have developed standout capabilities are going to get acquired and become part of a larger portfiolio. 

Reprints
Reprints can be purchased through Constellation Research, Inc. To request official reprints in PDF format, please contact Sales.